A Microsoft patent application for a ‘virtual reality floor mat’ describes a special mat which defines a virtual playspace and can be used for tracking.

Spotted by Twitter user WalkingCat, a Microsoft patent application published last week describes a “virtual reality floor mat activity region,” a special mat that would both define the virtual playspace and be used for tracking.

The artwork in the patent clearly suggests a living room environment where the VR experience is powered by a game console. When it comes to VR on Xbox, Microsoft has had a confused relationship; the company had announced way back in 2016 that VR support was coming to Xbox One X but then suddenly backpedaled on that decision. It has remained mum on the subject ever since, even in the face of the next-gen Xbox ‘Project Scarlett’ which was revealed earlier this year.

Next-gen Xbox Revealed with No VR in Sight, Sony Forges Ahead with PSVR on PS5

Microsoft’s virtual reality mat patent application is a curious one. With inside-out tracking increasingly becoming the norm on VR headsets (including Microsofts own Windows VR headsets which were among the first to market with the tech), it’s not immediately clear why a special mat would be necessary, but the patent application reveals some interesting possibilities.

Image courtesy Microsoft

First, the application describes that the mat would have active or passive markers on it which would make its location known to headsets. While inside-out tracked headsets need to have their playspace boundary defined, having a mat in your home would be a ‘set it and forget it’ approach to defining the playspace. Not only would this eliminate occasions where a system without such a mat might forget the boundary and require the user to reconfigure it, the mat-defined boundary would be accessible to any headset without per-device configuration, similar to the way that Valve’s SteamVR Tracking system can be used by any device that comes into the area without any pairing or configuration with the tracking system.

With the mat you could even drop inside-out tracking altogether; the mat could function as an outside-in tracking system (where the headset looks for only the markers on the mat and nothing else). This would reduce the compute requirements compared to inside-out tracking, which in turn could reduce the size and cost of a compatible headset and other devices. For instance, you could create a much less expensive controller which is tracked with the mat markers rather than its own on-board compute.

The mat would also facilitate a static and shared coordinate system which would be useful for multi-user experiences where two or more players are using a headset at the same time in the same space. This could also extend to AR devices like HoloLens, and even the tracking of non-immersive devices like smartphones or tablets, making possible various hybrid experiences.

And finally, there’s the human affordance factor. A mat on the floor makes clear to everyone else in the room where they shouldn’t stand if they don’t want to risk a haymaker to the noggin when someone is playing a VR boxing game or similar.

The patent application also considers a version of the mat which would be based on tiles which could be configured into novel shapes and expanded as needed by users.

Image courtesy Microsoft

Much of the functionality described in the patent sounds very similar to the approach of Antilatency which has developed a tile-based outside-in tracking system with minimal compute requirements which can functional well with just a single camera on the headset. The system is inherently expandable, doesn’t require per-device configuration, supports arbitrary numbers of users, and can be configured into novel shapes. The company is already selling dev kits for the system which is aimed at enterprise and commercial use-cases.

Image courtesy Antilatency

Microsoft’s virtual reality mat patent application is attributed to inventors Julia Schwarz, Principal Software Engineer on HoloLens, and Jason Michael Ray, Software Engineer on HoloLens. The application was filed in April 2018 and published on October 3rd, 2019, though it has not yet been granted.

As ever, it’s worth pointing out that large companies file hundreds if not thousands of patents each year as a strategic exercise; there’s no telling whether or not something described in a patent will manifest in a product. Without additional evidence, patents are best interpreted as a glimpse into what a company is thinking rather than what it is doing.

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. See here for more information.

  • dota

    Why not a vibration suit

  • Roman Vdovchenko

    Hi Ben. Could you please describe your thoughts about the “inside-out” and “outside-in” approach? Technically, Lighthouse and Antilatency are “inside-out” based systems because trackers should be placed on an object instead of “outside-in” when markers should be placed on a tracking target. We can use the first demo of Oculus Quest as an example. They use “markers” taped to the floor to determine position with maximum quality and precision. We use the same way but not with SLAM.

    • benz145

      Sure. The vernacular in the VR industry positions ‘outside-in’ as any tracking system that requires external trackers/beacons/sensors, while ‘inside-out’ is understood to mean a system which doesn’t require any external trackers/beacons/sensors.

      I’m well aware of the definition which also considers which device is actually doing the ‘seeing’, but as far as the VR end-user is concerned, that’s technical minutia which doesn’t really impact their experience one way or the other. That’s why the definition of inside-out and outside-in are used in the VR space are as they are. It would be nice if there were different terms for it so that we would have greater specificity, but alas this is how it happened.

      • Ray

        Why not use “self tracking headset” and headsets that use “assisted tracking”? Where assisted tracking can be outside-in using outside processing, such as the Rift cameras, or inside-out using markers such as black and white images and lighthouses.

  • MosBen

    This immediately made me think of Tilt 5 and their play mat.

  • gothicvillas

    In theory these mats could be in all possible designs. Interesting.

  • More crap for future patent trolls. There’s not a thing there that has to do with the future of VR. A tracking system that fails as soon as you look up? Junk cluttering the floor? All of the inside-out tracking systems struggle to track the hands controllers, not the headset itself. Nothing about this seems useful.

    Figure 170 and 172 seem to be the most important: the plate of nachos. Now if they are patenting nacho’s we might have reason to be concerned.

    • Jistuce

      Honestly, I find the mat brilliant as an indicator inside meatspace of where cyberspace resides.

      It simplifies setup for the user(no tracing out play areas with controllers, mounting sensors, etc) and takes things back to the PS1 and Dance Dance Revolution floor mats(throw it down and stomp around), and at the same time it produces a clear warning to everyone else. Everyone knows “don’t put your chairs here, that’s the VR area”, “don’t stack your books here, that’s the VR area”, “don’t step in here without warning the guy with the headset, that’s the VR area”.

      One assumes any future headset intended for use with this mat would have downward-facing cameras, much like the modern Windows Mixed Reality, Quest, Rift S, and Cosmos do.

      • In that case just buy a garish rug and a Quest.

        • Jistuce

          Also a viable option.

          • If only my dog understood this I would not have to keep half my brain in the real world lol

          • Jistuce

            Oh man, I have punched my dog so many times during Beat Saber.
            She doesn’t so much ignore VR as she does actively interfere with it.
            High action? Run right into the flailing arms, yelp in surprise when noggin is clobbered. Low action? Walk up to a stationary target, lay down as close to feet as possible without touching them(she isn’t NEAR this attached when I’m NOT blinded to the real world).

    • Nick Wallace

      I don’t think you could be more wrong, the whole “Look up” part is easily fixed by having cameras have a wide FOV and aim downwards similar to the Cosmos and such. AntiLatency is using similar tech with supposedly great results so far so definitely not outside the realm of possibilities.

  • Dan DeMontmorency

    This is interesting wonder if AntVR Cyclops infrared carpet tiles is related.

  • I don’t think this will ever become used at home. Maybe in LBVR